def main(unused_argv): """Saves bundle or runs generator based on flags.""" tf.logging.set_verbosity(FLAGS.log) bundle = get_bundle() config_id = bundle.generator_details.id if bundle else FLAGS.config config = pianoroll_rnn_nade_model.default_configs[config_id] config.hparams.parse(FLAGS.hparams) # Having too large of a batch size will slow generation down unnecessarily. config.hparams.batch_size = min( config.hparams.batch_size, FLAGS.beam_size * FLAGS.branch_factor) generator = PianorollRnnNadeSequenceGenerator( model=pianoroll_rnn_nade_model.PianorollRnnNadeModel(config), details=config.details, steps_per_quarter=config.steps_per_quarter, checkpoint=get_checkpoint(), bundle=bundle) if FLAGS.save_generator_bundle: bundle_filename = os.path.expanduser(FLAGS.bundle_file) if FLAGS.bundle_description is None: tf.logging.warning('No bundle description provided.') tf.logging.info('Saving generator bundle to %s', bundle_filename) generator.create_bundle_file(bundle_filename, FLAGS.bundle_description) else: run_with_flags(generator)
def create_sequence_generator(config, **kwargs): return PianorollRnnNadeSequenceGenerator( pianoroll_rnn_nade_model.PianorollRnnNadeModel(config), config.details, steps_per_quarter=config.steps_per_quarter, **kwargs)
pass message = { 'status': 400, 'message': "File not found", } resp = jsonify(message) resp.status_code = 400 return resp if __name__ == '__main__': bundle_file = 'pretrained/pianoroll_rnn_nade.mag' with tf.Session(): tf.logging.set_verbosity(log) bundle = get_bundle() config_id = bundle.generator_details.id config = pianoroll_rnn_nade_model.default_configs[config_id] config.hparams.parse(hparams) config.hparams.batch_size = min(config.hparams.batch_size, beam_size * branch_factor) generator = PianorollRnnNadeSequenceGenerator( model=pianoroll_rnn_nade_model.PianorollRnnNadeModel(config), details=config.details, steps_per_quarter=config.steps_per_quarter, checkpoint=None, bundle=bundle) app.run(debug=True, host='0.0.0.0', use_reloader=False)